search

bamos / dcgan-completion.tensorflow

Image Completion with Deep Learning in TensorFlow
Other
1.31k stars 388 forks source link

there is nothing in the folder “sample” after 25 epoch #70

Open DsJay opened 6 years ago

DsJay commented 6 years ago

after run this command ./train-dcgan.py --dataset ./data/your-dataset/aligned --epoch 20 there is nothing in the folder “sample” as expected

DsJay commented 6 years ago

I have already solved this problem thanks

ChaoWei0606 commented 6 years ago

@DsJay ,hello,Can you tell me how you train?after I delete original checkpoint directory,I run this command ./train-dcgan.py --dataset ./data/your-dataset/aligned --epoch 20.The computer display "An existing model was not fount in the checkpoint directory. Initializing a new one"I use my own training set.

DsJay commented 6 years ago

@ChaoWei0606 This sentence is just a comment,training should already be in progress when you see that.

ChaoWei0606 commented 6 years ago

@DsJay Thank you for your reply.You're right. The problem is that I don't have enough training samples.

ChaoWei0606 commented 6 years ago

@DsJay I have a question.why is the picture of the 'outputImage/completed' directory is similar to the picture of marked?

I mean it doesn't change anything.I use my own training set.

Do I need to do anything else?

DsJay commented 6 years ago

@ChaoWei0606 maybe you need a larger dataset